The agent assembled a comprehensive explanation of Marloo by working around the site's JavaScript-heavy rendering: direct fetches of marloo.com pages (steps 1, 3–5, 10) returned only boilerplate HTML and did not contain usable content, forcing the agent to rely entirely on web search results (steps 7, 8, 12, 13, 15) to extract pricing, features, positioning, and competitive context. The site itself is poorly navigable for machine agents due to client-side rendering, but the agent successfully answered all four dimensions of the task (what, who, pricing, differentiation) by synthesizing information from external sources.

• ›Direct page fetches to marloo.com/pricing, marloo.com/features, marloo.com/about, and marloo.com/us/pricing all returned 200 status but delivered only Next.js boilerplate HTML with no semantic content—the actual page content is rendered client-side and was completely inaccessible to the agent.
• ›All substantive content in the final answer (pricing tiers: Free + $149/user/month Pro; features: meeting notes, compliance docs, form completion, 'Ask Marloo'; competitive positioning; funding details: $12.7M Blackbird-led round) came from web search results, not from marloo.com itself. The agent cited fintech.global, tech.eu, thenextweb.com, and TechFundingNews as primary sources—none of which are the vendor's own site.
• ›The site's failure to publish structured or plaintext content meant the agent had to perform multiple searches with different query strategies (steps 7, 8, 12, 13, 15) to triangulate missing information—indicating poor machine-readability and requiring search as a workaround rather than site exploration.
• ›The agent explicitly flagged three key pieces of information it could not retrieve: enterprise pricing specifics, depth of integrations with CRM systems, and segment-specific feature details for insurance/mortgage modules—all genuine gaps in what the site publicly renders to agents.